The Algorithm Method

Making news decisions in a clickocracy

Journalists relate to their audiences differently in the age of online news, according to C. W. Anderson, in recent articles in Journalism and the International Journal of Communication. Both articles are based on research Anderson conducted in Philadelphia newsrooms. (Anderson is a professor of media culture at the College of Staten Island and blogs for NiemanLab; he received his PhD at Columbia’s journalism school where—full disclosure—he studied with Michael Schudson.)

According to Anderson’s research, the journalist-audience relationship has changed in part because it’s now easier to comment on news stories in a fast and public way. Letters to the editor can be e-mailed rather than snail-mailed. Online letters stand a much better chance of being published, and with less editing, than the small percentage selected for print. And while letters to the editor are often hidden away in the op-ed section, where reporters can ignore them, online comments tend to be affixed right to the end of offending articles.

And reporters are not thrilled. “Philadelphia is really full of a bunch of boorish jerks,” one reporter told Anderson after perusing comments on an article. That reporters often dismissed the comments as the work of “losers,” surely unrepresentative of the news audience as a whole, squares with earlier newsroom studies. What’s different now, Anderson contends, is that reporters aren’t surprised when they get audience feedback—they expect it, even if they often aren’t happy with it.

What’s also different is that audiences now provide feedback unintentionally through online metrics (the running tally of which articles get clicked on the most). Reporters—who fear that a lack of clicks could cost them their jobs—watch these tallies, as do editors and publishers, because higher metrics mean higher online ad revenues. To that end, news organizations have enlisted the assistance of online-tracking software and programmer analysts to monitor the ongoing flow of data. Despite substantial doubt about the validity of click data (see Lucas Graves and John Kelly, “Confusion Online: Faulty Metrics and the Future of Digital Journalism,” 2010, a report for The Tow Center for Digital Journalism), metrics have come to occupy the imagination of journalists and their organizations.

Still, if audience feedback, albeit in an aggregated click form, plays a bigger role in news decisions than it used to, could that mean that journalism is becoming more democratic? It depends on what you mean by “democratic.” Anderson examines three types of “outsider” media, each of which relate to the audience differently and each of which, he suggests, encapsulates a different conception of democracy. Those in “algorithmic” media relate to their audience in an aggregated, big-data kind of way. In his prime example of that, Demand Media, news outlets generate articles to suit what people search for online. “How to Build a Tractor Plow” and “How to Cut Egg Shells With a Laser Beam” are a couple of market holes waiting to be filled with articles from Demand Media writers.

But is this democracy? Of a mechanized, marketized sort, yes. This “democracy,” with its “algorithmic audience,” is worlds away from the public-journalism movement (Anderson’s second example of outsider media), which envisioned a sort of journalist-mediated giant town hall, with a “deliberative audience.” It’s also vastly different from the participatory form of democracy embraced by citizen journalism outlets like Indymedia (Anderson’s third example), which encourage their audiences to “be the media.” To be sure, algorithmic media also require audience participation—but only in an automated, aggregated clickocracy. There’s no opportunity for algorithmic audiences to explain why they clicked, whether they’re glad they did, or whether they’d click on something similar in the future.

How might algorithmic media influence public life? If algorithms come to dictate news decisions, how does that change what we read, and what sort of democracy we might have or want to have? Some media executives told Anderson that algorithmic journalism empowers audiences. Keep several grains of salt handy for that one! Perhaps the real power rests with journogeeks—the reporter-programmers who create the algorithms, and whose analysis of the resulting data drives news decisions down uncharted paths.

Has America ever needed a media watchdog more than now? Help us by joining CJR today.

Michael Schudson and Katherine Fink are contributors to CJR.